SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Declaration of Interests
  8. References

Background

Noroviruses (NoVs) are the most common cause of epidemic gastroenteritis, responsible for at least 50% of all gastroenteritis outbreaks worldwide and were recently identified as a leading cause of travelers' diarrhea (TD) in US and European travelers to Mexico, Guatemala, and India.

Methods

Serum and diarrheic stool samples were collected from 75 US student travelers to Cuernavaca, Mexico, who developed TD. NoV RNA was detected in acute diarrheic stool samples using reverse transcription-polymerase chain reaction (RT-PCR). Serology assays were performed using GI.1 Norwalk virus (NV) and GII.4 Houston virus (HOV) virus-like particles (VLPs) to measure serum levels of immunoglobulin A (IgA) and IgG by dissociation-enhanced lanthanide fluorescent immunoassay (DELFIA); serum IgM was measured by capture enzyme-linked immunosorbent assay (ELISA), and the 50% antibody-blocking titer (BT50) was determined by a carbohydrate-blocking assay.

Results

NoV infection was identified in 12 (16%; 9 GI-NoV and 3 GII-NoV) of 75 travelers by either RT-PCR or fourfold or more rise in antibody titer. Significantly more individuals had detectable preexisting IgA antibodies against HOV (62/75, 83%) than against NV (49/75, 65%) (p = 0.025) VLPs. A significant difference was observed between NV- and HOV-specific preexisting IgA antibody levels (p = 0.0037), IgG (p = 0.003), and BT50 (p = <0.0001). None of the NoV-infected TD travelers had BT50 > 200, a level that has been described previously as a possible correlate of protection.

Conclusions

We found that GI-NoVs are commonly associated with TD cases identified in US adults traveling to Mexico, and seroprevalence rates and geometric mean antibody levels to a GI-NoV were lower than to a GII-NoV strain.

Noroviruses (NoVs) are the most common cause of epidemic gastroenteritis, responsible for at least 50% of all gastroenteritis outbreaks worldwide, and a major cause of food-borne illness.[1] NoVs are a diverse group of viruses in the family Caliciviridae, and they are currently classified into six genogroups (GI–GVI) and further subdivided into multiple genotypes.[2] Strains that infect humans are found in GI, GII, and GIV, with GII.4 strains notably causing over 80% of NoV infections worldwide.[3] NoVs are highly infectious, very stable, and remarkably transmissible.[4, 5] Besides being a leading cause of epidemic acute gastroenteritis and an important cause of hospital acquired and sporadic infection, NoVs have also been described as a common etiology of travelers' diarrhea (TD).[6, 7] Interestingly, TD studies over time and in different locations have reported a marked diversity of GI- and GII-NoVs among cases.[6, 8-10] However, an almost equivalent number of NoV-associated travelers' diarrhea (NoV-TD) cases are associated with strains from both genogroups,[6] which contrasts with the predominance of GII.4 viruses being associated with outbreak settings and sporadic cases worldwide.[11, 12]

The role of the antibodies elicited after a NoV infection, particularly those of specific antibody classes, has been difficult to discern as in some studies individuals with high prechallenge enzyme-linked immunosorbent assay (ELISA) titers were more likely to become infected than individuals with low preexisting titers.[13, 14] More recently, the presence of antibodies that block the binding of NoV virus-like particles (VLPs) to histo-blood group antigens (HBGAs) prior to challenge or inhibit VLP binding to red blood cells have correlated with protection against clinical NoV gastroenteritis.[15-17]

The aim of this study was to investigate the prevalence of NoV genogroup-specific antibodies in a group of US travelers to Mexico and the occurrence of natural NoV infection during travel. Recombinant GI [Norwalk virus (NV)] and GII [Houston virus (HOV)] VLPs were used to determine the role of preexisting serum antibodies [immunoglobulin A (IgA), IgG, IgM and HBGA-blocking antibodies] in protection against NoV-TD.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Declaration of Interests
  8. References

Study Design

US student travelers (≥18 years of age) to schools in Cuernavaca, Mexico, during the summer of 2004, were invited to participate in a prospective study for the assessment of the etiology of TD and correlates of protection against the disease. Owing to the initial design of the study, only those travelers presenting TD symptoms provided stool specimens when ill. TD was defined as the passage of three or more unformed stools within a 24-hour period associated with one or more of the following symptoms: abdominal pain or cramps, flatulence, nausea, vomiting, fever [temperature, ≥100°F (≥37.8°C)], blood and/or mucus in the stool, or tenesmus. Exclusion criteria involved having other clinically important underlying illness, diarrhea for ≥72 hours, on current treatment against diarrheal pathogens, being pregnant, and breast-feeding.

A blood sample was drawn from all study participants within ≤48 hours of arrival in Mexico (referred to as preexisting) and at another time point ≤ 48 hours prior to departure (referred to as predeparture). During the course of stay in Mexico, a stool sample was obtained within 48 hours of onset of diarrhea. All samples were submitted to the laboratory at the Center of Infectious Diseases, University of Texas School of Public Health in Houston, Texas for analysis. Stool samples were screened for common enteric bacterial and parasitic pathogens using previously published methods.[18] Where sufficient stool was available, the samples were tested for NoV using methods previously reported.[6]

Definition of NoV-TD

A case was considered to be NoV-TD if the clinical description matched the definition of TD and NoV was detected in the stool sample by polymerase chain reaction (PCR) and/or there was a fourfold or more increase in antibody titer (HBGA-blocking antibodies) or concentration (IgA and IgG) between preexisting and predeparture serum samples using any of the serological assays described below.

Detection and Characterization of NoVs

Detection of NoVs was performed as described previously.[6] Briefly, viral RNA was extracted from 200 μL of a 10% stool suspension (in 0.01 mol/L phosphate-buffered saline) with a QIAamp viral RNA kit (Qiagen) according to the manufacturer's instructions, and the RNA was stored at −80°C until used. The reverse transcription (RT) reaction was carried out in 50 μL of a reaction mixture containing 75 pmol of random hexamers (Thermo Fisher Scientific, Waltham, MA, USA), 5 U of avian myeloblastosis virus reverse transcriptase (AMV-RT) (Life Sciences), 1× AMV-RT buffer (Life Sciences, St. Petersburg, FL, USA), 10 mmol/L of dNTP mix, 10 U of Protector RNase Inhibitor (Promega, Madison, WI, USA), and 10 μL of RNA extract. The reaction was performed for 2 hours at 43°C, and the enzyme was inactivated at 75°C for 5 minutes. The resulting cDNA was stored at −20°C until used in PCR assays. A set of degenerate primers[19] was used for initial detection of NoVs and a genogrouping PCR assay was performed on positive samples using primers G1SKF, G1SKR, G2SKF, and G2SKR.[20] All PCR products were separated by 2% agarose gel electrophoresis and visualized under ultraviolet (UV) lamp after ethidium bromide staining. Pathogens other than NoVs were previously sought in this population and reported elsewhere.[21]

Serology Assays

Anti-norovirus IgA and IgG were determined by dissociation-enhanced lanthanide fluorescent immunoassay (DELFIA) and IgM by a capture ELISA as described elsewhere,[22] with an adaptation to include detection of specific antibodies against HOV-VLPs as well as against NV-VLPs. For serum IgA and IgG detection, Ig concentrations (µg/mL) were calculated by interpolation from standard curves generated using purified IgG or IgA. For the serum IgM assay, a test sample was considered positive if the optical density (OD) value was above the average of blanks plus three-times the standard deviation of all the blanks. The blocking assay was carried out as described previously,[17] and was also adapted for testing HOV-VLPs using H-type 1 synthetic carbohydrate. The 50% blocking titer (BT50) was defined as the titer at which the OD reading (after subtraction of the blank) was 50% of the OD of the positive control. A value of 12.5 was assigned to samples with BT50 below the limit of detection of the assay (<25).

Statistical Analyses

Wilcoxon signed rank test was used for the comparison of paired nonparametric data. McNemar's test was used to compare the proportion of NV and HOV-IgA and IgG positive travelers.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Declaration of Interests
  8. References

Study Population and Sample Collection

The study population consisted of 75 US students (age: 18–76 years, mean: 26 years; median: 22 years), of whom 86% were Caucasian, 84% of non-Hispanic ethnicity, and 54% were female. The average length of stay in Cuernavaca was 32 days (median: 37 days, range: 12–53 days). Paired (preexisting and predeparture) serum samples were collected from 75 travelers who experienced TD. Of these, diarrheal stool samples were collected from 56 travelers for enteropathogen studies. After initial testing for bacteria and parasites, sufficient stool was available from 37 subjects for screening for NoV. Of the travelers who provided both blood samples along with a stool sample, the average time between the onset of symptoms and the predeparture blood draw was 16 days (median: 13 days; range: 1–45 days).

Prevalence of NoV Among TD Patients

Ten of the 37 stool samples (27%) tested positive for NoV by reverse transcription-polymerase chain reaction (RT-PCR). Of these, seven samples (70%) were identified as GI-NoVs, and three (30%) were identified as GII-NoVs. Additionally, two travelers, from whom stool samples were not available for NoV detection, were diagnosed as cases of GI-NoV-TD based on a fourfold or more increase in serum antibody titers (>15-fold increase NV-IgA and more than fourfold increase in NV-IgG for both travelers). Overall, 12 (16%) of the 75 travelers were determined as having NoV-TD. Only one individual for whom a stool sample was available and NoV was detected (GI-NoV) had a fourfold or more increase in serum antibody titers. Among NoV-positive travelers, the average duration between onset of symptoms and the predeparture blood sample draw was 22 days (median: 23 days, range: 10–36 days).

Preexisting IgA

Among preexisting serum samples, 49 of 75 (65%) samples were found to have measurable levels of NV-IgA (≥1.6 µg/mL), whereas significantly more individuals (62/75, 83%) had HOV-specific IgA (p = 0.025) (Figure 1). The median concentration of preexisting serum IgA against NV-VLPs was 1.67 µg/mL (range: 1.13–11.91 µg/mL), and was also significantly lower than the median concentration against HOV, which was 2.04 µg/mL (range: 1.06–23.97) (p = 0.0037).

image

Figure 1. Comparison between preexisting serum antibody levels against Norwalk virus (NV) virus-like particle (VLP) and Houston virus (HOV) VLP. (A) Anti-NV and -HOV IgA levels, p = 0.0037; Wilcoxon signed rank test, cutoff: 1.4 µg/mL. (B) Anti-NV and -HOV IgG levels p = 0.003; Wilcoxon signed rank test, cutoff: 1.6 µg/mL. (C) Blocking titer (BT50) values of NV- and HOV-blocking antibodies, p = <0.0001; Wilcoxon signed rank test. Bars represent geometric means and dotted lines the cutoff value.

Download figure to PowerPoint

However, the level of preexisting IgA antibodies against NV or HOV did not correlate with protection against GI- or GII-NoV infection (Table 1). The frequency of development of GI-NoV diarrhea was similar among NV-IgA seropositive (12.2%) and seronegative (11.5%) individuals. Similar results were observed for HOV-specific IgA where all the three individuals infected with GII-NoVs had detectable levels of HOV-specific IgA.

Table 1. Distribution of total (N) and NoV-TD travelers by preexisting levels of IgA, IgG, and BT50 titers against NV- and HOV-VLPs
 NV-IgANV-IgGNV-BT50a
<1.6 μg/mL>1.6 μg/mL<1.4 μg/mL>1.4 μg/mL<25>25
N = 7526 (35)49 (65)075 (100)61 (89)14 (11)
NoV-TD4 (33)8 (67)012 (100)10 (83)2 (17)
GI-NoV360972
GII-NoV120330
 HOV-IgAHOV-IgGHOV-BT50a, b
<1.6 μg/mL>1.6 μg/mL<1.4 μg/mL>1.4 μg/mL<25>25
  1. Numbers shown are n (%).

  2. NoV-TD = noroviruses-associated travelers' diarrhea; Ig =immunoglobulin ; BT50 = blocking titer; NV = Norwalk virus; HOV = Houston virus; VLPs = virus-like particles.

  3. a

    For BT50, travelers are classified having titers above or below 25 (<25, seronegative and >25, seropositive), and for IgA (detection limit 1.6 μg/mL) and IgG (detection limit 1.4 μg/mL) are classified by seropositivity. The number in parenthesis indicates the percentage of NV-TD travelers in the specific group.

  4. b

    Two GI-NoV positive samples not included due to lack of sera.

N = 7523 (31)52 (69)075 (100)36 (49)37 (51)
NoV-TD1 (8)11 (92)0125 (50)5 (50)
GI-NoV180934
GII-NoV030321

Preexisting IgG

All the samples tested had measurable levels of IgG against NV and HOV-VLPs (Table 1). The median concentration of antibodies against NV-VLPs in preexisting serum samples was 2.19 µg/mL (range: 1.70–23.82 µg/mL), and that against HOV-VLPs was 3.33 µg/mL (range: 1.81–20.83 µg/mL). Overall, travelers had higher preexisting concentration of serum IgG against HOV-VLPs than to NV-VLPs (p = 0.003, Figure 1). However, the presence of preexisting antibodies did not correlate with protection against GI- or GII-NoV infection.

Immunoglobulin M

Preexisting levels of IgM were detected in one serum sample using NV-VLPs. This individual had neither detectable levels of NoV in stool nor an increase in IgA or IgG antibody titer. The predeparture serum sample of this individual remained positive for IgM anti-NV-VLP. None of the serum samples tested with HOV-VLPs had detectable levels of serum IgM (data not shown).

HBGA-Blocking Antibodies

Of the 75 serum samples collected upon enrollment that were tested for H-type 1 blocking antibodies, only 14 (19%) had BT50 > 25 against NV-VLP compared with 37 (51%) against HOV (p < 0.0001). The median BT50 against NV was 12.5 (range: 12.5–67.2) compared with 27.1 (range: 12.5–3091.7) for HOV (p = <0.0001).

There was no difference between preexisting BT50 of infected and uninfected travelers for NV infection [infected: median NV-BT50 12.5 (range: 12.5–32.80); uninfected: median NV-BT50 12.5 (range: 12.5–67.20), p = 0.714; Table 1]. None of the travelers with NoV-TD had BT50 > 200, suggesting that the overall seroprevalence of the putatively protecting antibody levels was low among this group.

NoV Single Infection and NoV Coinfections

Of the NoV positive travelers, the presence of copathogens was observed in nine (75%) of the samples. The most common copathogen found was enterotoxigenic Escherichia coli (ETEC), which was present in seven (58%) of NoV-positive samples.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Declaration of Interests
  8. References

The prevalence of NoV-associated TD found in this study was 16%, similar to what has been reported previously for comparable populations traveling to Mexico (circa 15%).[6, 9, 10] However, this result is an underestimate of the prevalence of NoV-TD in international travelers as we only studied diarrhea and not gastroenteritis with vomiting, a much more common presentation of illness caused by this group of viruses, and only half of the stool samples collected were available for testing. Similar to previous reports, GI-associated NoV-TD cases were more prevalent (75%) than GII-NoVs.[8, 9] This result supports previous studies confirming that NoVs are a commonly identified pathogen of TD in US travelers to Mexico.

To detect baseline levels of antibodies against NoVs among travelers, we chose to use VLPs from two different strains of NoV-VLPs (GI.1-NV and GII.4-HOV) that represent the predominant genotype within genogroups I and II, respectively. In this context, a significantly greater number of travelers had preexisting IgA antibodies to HOV-VLPs than to NV-VLPs (p = 0.025) and in higher titers (p = 0.0037). The higher seroprevalence rates for GII.4 may be due to the predominance of these strains worldwide.[3] Although a confirmed history of previous NoV infection was not available for any of the study participants, serum samples from all study participants had IgG levels against NV- and HOV-VLPs above the assay cut-off, suggesting at least one previous exposure to the virus. The presence of preexisting antibodies against NoVs was not associated with protection specifically from GI or GII NoVs; however, the small number of test samples and the lack of testing of travelers who did not experience TD, limit this conclusion. Nonetheless, previous studies have demonstrated that a prior NoV infection does not confer long-term homologous protection against subsequent infections and disease.[13, 14]

Previous results from our laboratory and others suggest that antibodies capable of inhibiting the binding of VLPs to HBGAs or red blood cells have correlated with protection against disease.[15-17, 23] The frequency and the level of blocking antibodies found among travelers against HOV-VLP were higher than that against NV-VLP (p = <0.0001), in agreement with the serum IgA and IgG results. Of the 12 travelers infected with NoVs, 9 were considered seronegative for blocking antibodies (BT50 < 25) and 3 others had titers below the suggested threshold for protection against infection and viral illness (BT50 < 200).[23] Future comprehensive studies are needed to confirm cross-reactivity of protection conferred by blocking antibodies.

Only 3 of 12 infected individuals had a fourfold or more rise in antibody titers following infection (data not shown). However, this difference was not attributed to the length of time between samples for both groups as individuals who showed a fourfold or more rise in antibody titers following infection had an average of 33 days between sample collection, similar to that of 29 days for individuals who did not have a significant sero-response. Nevertheless, these results could underestimate the antibody response in this population as the serological assays were carried out with VLPs representative of GI and GII viruses that may not have been homologous to the virus detected in stool, thus decreasing the sensitivity of the assay. For future studies, homologous VLPs for determining sero-responses to NoV infection should be used because of the high capsid sequence diversity observed in this group of viruses. Nonetheless, the DELFIA assay used in this study has previously been shown to be highly specific when compared with antigen detection in stool.[22] The small quantity of stool samples was a limiting factor in carrying out genotyping assays in this study.

A high level of coinfections was observed among travelers with NoV-TD. This observation is in line with our previous studies in travelers where rates of coinfections are found to be over 60%.[7] TD is characterized as multifactorial disease caused by a wide range of enteric pathogens including bacteria, viruses, and parasites.[24] Furthermore, the ratio of pathogen coinfection in ill travelers is significant among the total number of cases. Other pathogens in addition to viruses must be sought for when defining the etiology of TD. However, it should be noted that the presence of copathogens did not appear to interfere with the stimulation of the immune response to NoVs as the travelers who showed a fourfold or more rise in antibody had multiple enteric pathogens detected in their stools.

In conclusion, we provide more evidence that NoVs are associated with the development of TD among US travelers to Mexico. Moreover, not only fewer travelers have antibodies against a GI-VLP compared with a GII-VLP, but also the amount of specific antibodies detected was lower for a GI-VLP than a GII-VLP. These results could partially explain the continuous presence of GI-NoV infections in travelers to these locations.[8, 9] Altogether, international travelers to developing regions of the world represent a unique and traceable population with relatively low prevalence of blocking antibodies and high rates of NoV endemicity, making them valuable for evaluation of NoV vaccines in the pipeline for development.[25]

Acknowledgment

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Declaration of Interests
  8. References

Research sponsorship was the following: NIH (PO1-AI057788; P30-DK56338) and USDA (2011–68003.30395).

Declaration of Interests

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Declaration of Interests
  8. References

R. L. A. has received research support from and is a consultant to Takeda Vaccines (Montana), Inc. M. K. E. received support from NIH/NIAID to conduct the study. She is an inventor on patents related to cloning of the Norwalk virus genome and is a consultant to Takeda Vaccines (Montana), Inc.

The other authors state they have no conflicts of interest to declare.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Declaration of Interests
  8. References